This proposal is for a competing continuation of a study of determinants of HIV risk among Puerto Rican injection drug users (IDUs) and crack smokers in New York and in Puerto Rico. Based on data from the parent grant, this proposal was developed to test the efficacy of an intervention (The Bienvenidos Project) developed to reduce risk behaviors among Puerto Rican migrant drug users (defined as those who previously used heroin or cocaine in Puerto Rico and now use drugs in the NY area). Our research found that among Puerto Rican IDUs in NY, those who were migrants (39% of our IDU sample) were riskier (reported higher levels of shooting gallery use and sharing injection equipment). Our data also showed that these riskier practices were related to the higher levels of risk behaviors found in Puerto Rico. The Bienvenidos Project incorporates interventions at the social (Peer) level and the drug treatment program level. The study will use a group randomized trial design, and will be based in 4 pairs of MMTP clinics with substantial proportions of Puerto Rican clients. Within each pair, one clinic will be randomly assigned to an Intervention Condition, which will (a) provide joint training of staff and Peer Outreach Workers regarding the higher risk behaviors of Puerto Rican migrant drug users, and (b) train Peers in skills regarding outreach and risk reduction with migrants. In the Control Condition, Peers will receive a brief non-HIV related training and staff will receive delayed training. Assessments of the intervention's impact will be conducted at the clinic level, and with staff (n= 120), peers (n=120) and migrants (n=560). Qualitative data collection will provide contextual information to enhance understanding of the intervention's impact. Using hierarchical linear modeling as the primary statistical method, hypotheses to be tested include: (a) MMTP clinics in the Intervention Conditions will enhance their services to address needs of migrants; (b) MMTP staff in Intervention clinics will increase their knowledge about migrants' HIV risks and their communication with clients and staff regarding migrants;(c) Peers will reduce their drug and sex-related risk behaviors; and (d) Migrants recruited from the Intervention clinics' communities where Peers do outreach, will reduce their drug and sex-related risk behaviors. The results of this study will have implications for conducting interventions with migrant drug users, and will enhance our knowledge regarding multi-level interventions to reduce HIV risk.